2021-05-18 IfcAlignmentVerticalSegment - is the expressed function meant to be continuous or not

Ref IFC Rail schema 4x3 8.7.3.4.1

The text should be more explicit on having an elevation coordinate (the "Z") which is continuous. As the text is written now, the text only states that the continuity requirement affects the domain (the segments are contiguous end-to-start along the distance-along-axis), not the codomain.

Lacking statement on : `Segment[i].Z(end) == ? Segment[i+1].Z(start)` (co-domain continuity).

This constraint should be encoded in EXPRESS.

"End-to-start" means: `Segment[i].StartDistAlong + Segment[i].HorizontalLength = Segment[i+1].StartDistAlong` - **ok**
"Not enforced to be tangential" means: `NOT Segment[i].EndGradient EQUALS Segment[i+1].StartGradient` - **ok**

## Example from Storyline 06 RFI alignment

The IFC file from ACCA was truncated to lower precision (which is not an error in itself), leaving start- and end gradients to be slightly inconsistent with a circular curve of the given HorizontalLength and CurveRadius. If taken at face value, the derived segment end height `Segment[i].Z(end)` coordinate will not be the same as the following segment's start height `Segment[i+1].Z(start)`. But if the importing software __assumes__ that Z should be contiuous as well, then the original precision can be reconstructed.